Design of Oil and Gas Composite Pipes for Energy Production

Available online at www.sciencedirect.com Available online at www.sciencedirect.com Available online at www.sciencedirect.com Available online at www.sciencedirect.com Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com Available

ScienceDirect ScienceDirect ScienceDirect ScienceDirect

Energy Procedia 106 (00 2016 ) 1 000–000 Energy Procedia (2019) Energy Procedia 106 2016 ) 1000–000 Energy Procedia (2019) 000–000 Energy Procedia 162 (2019) 146–155 Energy Procedia 00 (00 (2017)

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

Special Issue ontheEmerging and Energy: Generation Automation The 1stst edition Energy Economics Iberian Conference took place on 4-5and February Special Issueof Emerging and Renewable Renewable Energy: Generation Automation The 1 edition ofon the Energy Economics Iberian Conference took place on 4-5and February

2016 in Lisbon, Portugal. The event was organized by the Portuguese Association for 2016 inof Lisbon, The event was organized by the Portuguese AssociationProduction for Design Oil and Gas Composite Pipes for Energy Energyof Economics (APEEN) in close collaboration with its Spanish (AEEE). Design OilPortugal. and Gas Composite Pipes forcounterpart Energy Production Energy Economics (APEEN) in close collaboration with its Spanish counterpart (AEEE).

Both are respectively, the Portuguese and theAli Spanish IAEE – International Tamer Sebaey Both areThe respectively, the Portuguese and theAli Spanish affiliates ofHeating IAEE – International 15th International Symposium on affiliates Districtof and Cooling Tamer Sebaey Association for Energy Economics. Association for Energy Economics. Mechanical Design and Production Dept., Faculty of Engineering, Zagazig University, P.O. Box 44519, Zagazig, Sharkia, Egypt Mechanical Design and Production Dept., Faculty of Engineering, Zagazig University, P.O. Box 44519, Zagazig, Sharkia, Egypt The Conference had 155 participants from 17 countries and 73 papers were presented, The Conference had 155 participants from 17 countries and 73 papers were presented, from a total of 90 sent to be evaluated by the Scientific Committee, after the initial from a total of 90 sent to be evaluated by the Scientific Committee, after the initial submission of 142 abstracts. From those 73 presented papers, 22 were selected to be submission 142 abstracts. From those 73 presented papers, 22 were selected to be Abstract published inof issue of Energy Procedia. Abstract published in this this Procedia. a a,b,c issue of Energy a b c c Fiber reinforced composites *, pipes strength,and characteristics and high corrosion I. Andrić A.provide Pinaexcellent , experience P. Ferrão J. stiffness Fournier .,both B. for Lacarrière , O. and Leerosion Correresistance. Thecomposites Iberian energy market may valuable other Fiber reinforced pipes provide excellent strength and be stiffness characteristics and high European corrosion and erosion resistance. In addition, the to tailor market the strength and stiffness optimizing the winding angle gives the designer Thepossibility Iberian energy experience maycharacteristics be valuablebyboth for other European a In addition, the possibility toalso tailor the strength and stiffness characteristics by optimizing the winding angle gives thePortugal designer countries but worldwide, for -Latin American countries closely linked to Center for Innovation, Technology and Policy Instituto Superior Técnico, Av.properties Rovisco Pais 1, 1049-001 Lisbon, extra IN+ flexibility to design different pipe based onnamely theResearch different working conditions. These make them attractive for several countries but also worldwide, namely for Latin American countries closely linked to b extra flexibility to design different pipe based on the different working conditions. These properties make them attractive for several Veolia Recherche & Innovation, 291 Avenue Dreyfous Daniel, 78520 Limay, France Portugal and Spain by and applications including lightweight andculture, efficientlanguage equipment forbusiness. energy production applications. Portugal and Spain by culture, and c applications including lightweight and efficientlanguage equipment forbusiness. energy production4applications. Département Systèmes Énergétiques et Environnement IMT Atlantique, rue Alfred Kastler, 44300 Nantes, France In the current work, glass fiber reinforced plastic (GFRP) pipes were designed with four different winding angles. The pipes In the current glass fiber reinforced plastic (GFRP) pipes were designed with four winding angles. The pipes Bothwork, under the security perspective and under sustainability, thedifferent Iberian case were tested under pressure and of lowsupply velocity impact. Four different were manufactured winding with Bothinternal under the security of supply perspective and under designs sustainability, the Iberianusing case filament were tested under internal pressure and low velocity impact. Four different designs were manufactured using filament winding with can be viewed as a case study. It is still an energy island in Europe which increases winding angles of [±45/ ± 45/ ± 45], [±55/ ± 55/ ± 55], [±63/ ± 63/ ± 63], and [±63/ ± 45/ ± 55]. Each pipe has internal diameter canofbe viewed a case study. It± 55], is still an ± energy island Europe which increases winding angles [±45/ ± 45/as ± 45], [±55/ ± 55/ [±63/ 63/ 63], andin[±63/ ± 45/ ± 55]. Each pipe diameter to consumers, reduces competitiveness and±pipes creates vulnerability in terms of hastointernal of 110 mm, costs wall-thickness of 3.8 mm, and length of 450 mm. The were exposed to internal pressure determine to consumers, reduces competitiveness and pipes creates vulnerability in terms of to determine their ofAbstract 110 mm, costs wall-thickness of The 3.8 mm, andbill length of 450 consumers mm. The werehigher exposed to in internal pressure energy supply.impact energy of impact Iberian is 40% than France. A capacity was 56 their capacities and low velocity to assess their resistance. Under internal pressure, the maximum energy supply.impact The energy of impact Iberianresistance. consumers is 40% higher than in A capacity was 56 bars capacities and low velocity to assessbill their Under internal pressure, theFrance. maximum bars Iberian gas corridor is a particularly urgent infrastructure needed not the governed by matrix angles. All specimens failed in the same way –ofbut initial leakage, and recordedreal for the pipes with [±55]3 winding real Iberian gas corridor is aaddressed particularly urgent infrastructure needed but not the governed angles. in Allthe specimens failed in the same way –of initial leakage, by matrix and recorded for the pipes with [±55] District heating networks are commonly literature as one of the most effective solutions for decreasing the 3 winding cracking, which in the internal pressure. For the impact resistance, the combined orientations ([±63/ ± 45/ ± 55]) onlycauses one - atodrop correct this situation. cracking, which causes atodrop in the the internal pressure. For the impact resistance, the combinedwhich orientations ([±63/ ± 45/ the ± 55]) only one correct this situation. greenhouse gas emissions from building sector. These systems require high investments are returned through heat showed higher assessment, compared to the other pipes. This higher damage resistance can be justified by the mismatch angle

Assessing the feasibility of using the heat demand-outdoor temperature function for a long-term district heat demand forecast

showed higher compared to the otherand pipes. This higher damage resistance be justified the mismatch angle sales.ofDue to assessment, the changed climate conditions building renovation policies, heatcan demand in thebyfuture could decrease, effect the adjacent plies. Although Spain´s dependence on energy imports was still 70% in 2014, there was an effect of the adjacent plies. prolonging the investment return period. Although Spain´s dependence on energy imports was still 70% in 2014, there was an impressive reduction of 10% comparedofto 2009. Alsodemand Portugal reduced its energy The main cscope of this paper is to assess the feasibility using the heat – outdoor temperature function for heat demand impressive reduction of 10% to 2009. Also Portugal reduced its energy Copyright Authors. 2019 Elsevier Ltd. by All rights © 2019 The Published Elsevier Ltdcompared from 90% in reserved. 2005 to 71%(Portugal), in 2014.was In used both ascountries, incentives to is consisted of 665 c dependency Copyright  2019 Elsevier Ltd. All rights reserved. forecast. The district of Alvalade, located in Lisbon a case study. The district from 90% in 2005 to 71% in 2014. In both countries, incentives to Selection and anddependency peer-review under responsibility of the scientific committee of the 6th International Conference on Emerging and peer-reviewenergy under responsibility of the scientific of the SpecialInIssue onreal Emerging Renewable Energy: renewable generation were crucial tocommittee this performance. 2015 energyand Selection peer-review under responsibility of the scientific of the Special onreal Emerging Renewable Energy: buildingsand that varyGeneration in both construction period and typology. Three weather scenarios (low, medium, high) and three district Renewable Energy: and Automation, ICEREGA 2018. renewable energy generation were crucial tocommittee this performance. InIssue 2015 energyand Generation and Automation. generation from renewable sources represented respectively, fromobtained electricity Generation Automation. renovationand scenarios were developed (shallow, intermediate, deep). To estimate48,1% the error, heat demand values were generation from renewable sources represented respectively, 48,1% from electricity Keywords: gross generation plus importing balance in Portugal and 38,4% inand Spain. compared with results from a dynamic heat demand model, previously developed validated by the authors. Keywords: gross plus importing balance in Low Portugal and 38,4% in Spain. Pips; pipes;generation Glass reinforced internal pressure; velocity TheComposite results showed that when onlyplastics; weather change is considered, the impact margin of error could be acceptable for some applications Pips; Composite pipes; Glass reinforced plastics; internal pressure; Low velocity impact (the error inNotwithstanding, annual demand was lower than weather scenarios sector considered). However, after introducing renovation according to 20% IEA for theall power generation remains the largest Notwithstanding, according to IEA (depending the powerongeneration sector remains scenarios the largest scenarios, the error value in increased to 59.5% the andtotal renovation combination considered). CO2 emitter Portugalupwith 15.8 MtCO2 in 2013 orweather 35.2% of CO2 emissions. In CO2 emitter in Portugal with 15.8 MtCO2 in 2013 or 35.2% of total The value of increased average within range of 3.8% up toCO2 8%emissions. perMtCO2 decade,In theslope casecoefficient of Spain, the energyonsector is also thethelargest CO2 emitter with 87 inthat corresponds to the of of Spain, the hours energy sector is during also the largest CO2 emitter with 87onMtCO2 in decrease in the the case number heating ofsector 22-139h the heating season (depending the while combination of weather and 2013. The power generation accounted for 28.4% of total CO2 emissions, 1. Introduction 2013. Theconsidered). power generation sector accounted for 28.4% increased of total CO2 emissions,per while scenarios On the other hand, function intercept for 7.8-12.7% decade (depending on the 1.renovation Introduction other energy industries (including oil refining) accounted for a further 7.9%. other energy industries (including a further 7.9%. for the scenarios considered, and coupled scenarios). The values suggested couldoil berefining) used to accounted modify the for function parameters Polymeric composites, of resin (epoxy, polyester, PEEK, etc.) and reinforcing fibers (carbon, glass, improve the accuracy of heatcomposed demand estimations. Polymeric composites, composed of resin (epoxy, polyester, PEEK, etc.) and reinforcing fibers (carbon, glass,

To show attain high CO2 reduction goals, both haveand two alternatives in fact, aramide, etc.) specific strength and Portugal stiffness,and lowSpain density, high chemical––resistance. Fiber reinforced To show attain high CO2 reduction goals, both Portugal and Spain haveand two alternatives in fact, aramide, etc.) specificpolicies strength and stiffness, lowimplemented density, high chemical resistance. Fiber reinforced two complementary – which must be wisely but broadly: to not polymeric composites, as the by most important category of engineering composites, are employed only in high © 2017 The Authors. Published Elsevier Ltd. two complementary policies – which must be implemented wisely but broadly: to polymeric composites, as the most important category of engineering composites, are employed not only in high improve energy efficiency gains and applybut a have CO215th tax which revenues should be such technology industrial applications as aerospace, also penetrated low-tech industry as and sanitary Peer-reviewimprove under responsibility of thesuch Scientific Committee ofa The International Symposium on District Heating energy efficiency gains and apply CO2 tax which revenues should be technology allocated industrialtoapplications such as aerospace, but have also penetrated low-tech industry such as sanitary related innovation. Cooling. allocated to energy energy related innovation. ∗ Tamer A. Keywords: Heat demand; Climate change Tel.:Forecast; +2-011-560-23-516 ∗ Tamer A. Sebaey. Sebaey. Tel.: +2-011-560-23-516 E-mail address: [email protected] E-mail address: [email protected]

1876-6102Copyright © 2017 The Authors. Published by rights Elsevier Ltd. c 2019 1876-6102  Elsevier Ltd. All reserved. c 2019 Elsevier 1876-6102 Copyright  Ltd. All rights reserved.of The 15th International Symposium on District Heating and Cooling. Peer-review under responsibility of the Scientific Committee Selection and peer-review under ofcountries: the scientific committee of the Special Issue on Emerging and Renewable Energy: Generation IEA (2016), Energyresponsibility Policies of IEA Portugal 2016 review. OECD/IEA, 2016 1876-6102 © 2019 The Authors. Published by Elsevier Ltd. Selection and peer-review responsibility ofcountries: the scientific committee of the Special Issue on Emerging and Renewable Energy: Generation (2016),under Energy Policies of IEA 2016 review. OECD/IEA, and Automation. Selection and IEA peer-review under responsibility of the Portugal scientific committee of the 6th 2016 International Conference on Emerging and and Automation. Renewable Energy: Generation and Automation, ICEREGA 2018. 1876-6102 © 2016 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license 10.1016/j.egypro.2019.04.016 1876-6102 © 2016 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license http://creativecommons.org/licenses/by-nc-nd/4.0/). http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review by the scientific conference committee of EEIC | CIEE 2016 under responsibility of Guest Editors. Peer-review by the scientific conference committee of EEIC | CIEE 2016 under responsibility of Guest Editors. doi:10.1016/j.egypro.2016.12.099 doi:10.1016/j.egypro.2016.12.099

2

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

147

ware. Indeed, the market for polymeric composites now spans the full range of industry sectors, including transport (rail, road, air and sea), military, aerospace, municipality, energy production and transmission, civil and infra-structure, sports and leisure [1, 2]. In the pipeline industry, driven by the ever-increasing need for energy and water resources, the market is rapidly growing, with fiber reinforced pipes one of the key potential materials [3]. Composite pipes currently find applications in chemical industry, ducts, offshore, water supply and sewage systems [4]. Typically steel has been used in piping applications, which provides good performance, especially under heavy mechanical loading (e.g. high pressure, large pipe movement,etc.) [5, 6]. However, in aggressive environments, steel pipes undergo degradation because of internal or external corrosion, which can generate partial or total failure of the pipe [7, 8]. For this reason, several studies have focused on the search for new resistant and non-corrosive materials and claddings [9, 10]. Glass fiber reinforced plastic (GFRP) pipes represent an attractive alternative to steel pipelines subjected to severe internal or external environments in onshore or offshore applications due to its corrosion resistance properties, which reduces maintenance and costs and lengthens the lifetime of the pipe [11]. A service lifetime of 50 years is generally considered for civil engineering structures. GFRP pipes are also required to remain in service for 50 years as a long-term design constraint in accordance with international rules and regulations [12]. In addition to the high internal pressure capacity driven by the high strength and long lifetime, the low density of composite pipe results in reduction in construction and transportation costs [13]. Composite pipe body is usually made of three main components: an inner liner, a composite laminate, and an outer cover [14]. The liner, either metallic or polymer, serves mainly as a barrier against the inner fluid [15]. The laminate is the load bearing component. The outer cover is a protective layer from the external environment. The inner liner, laminate and cover are all bonded or fused to the laminate. Additional outer layers may be added as special purposes layers, e.g. as local wear protection or fire protection. Composite pipes under internal pressure are subjected to both hoop and axial stresses in the case of close-ends tubes with hoop-to-axial stress of two to one. For long open-ends, axial stress resulting from the internal pressure is zero. In addition, other stresses may result from the installation, weight, external pressure, etc. In designing composite pipes, the stress and failure analysis is generally performed without considering either the internal liner nor the outer cover, i.e. it is assumed they do not contribute to the resistance to deformation [14]. The design process of the composite lay-up may include fiber and matrix material selection, overall laminate thickness, the thickness of each lamina, and the fiber orientation of each individual layer. Designers should consider that internal pressure does have an effect on the measured mechanical properties of FRP pipes when being tested under pressure and different mechanical loadings [16]. For thin-walled cylindrical-pressure vessels with a ratio of applied hoop-to-axial stress of two to one, an optimum winding angle of 55◦ was noted. For thicker walled tubes, the optimum angle changed slightly [17]. The optimum winding angle also depends on the loading condition, [18]. Hamed et al. [19] showed that filament wound pipes should be wound at ±55◦ for biaxial pressure loading, ±75◦ for hoop pressure loading, while ±85◦ is suitable for biaxial pressure with axial compressive loading. Haftchenari et al. [20] tested Kevlar/epoxy composite tube under internal pressure at different temperatures. The authors used winding angles of ±25◦ , ±55◦ , and ±75◦ . The results showed that for the 55◦ angle, the hoop strain is a function of the temperature whereas, the hoop stress does not show any dependency. For the other winding angles, the seems to be independent of the working temperature. This result can be justified by the difference in the coefficient of thermal expansions of the three laminates. Rafiee and Mazhari [21] studied the effect of stacking sequence, fiber volume fraction and winding angle on the internal pressure capacity of composite pipes using finite-element modeling. Three winding angles were examined: 52.5◦ , 57.5◦ , and 60.19◦ . Similarly, the numerical analysis of Sulu and Temiz [22] showed that the internal pressure capacity of GFRP pipe is a function of stacking sequence as well as wall thickness. It is worth remarking that the model presented by Rafiee and Mazhari [21] considered more material and damage features whereas, the study in [22] examine a wider range of winding angles. Krishnan et al. [23] also studied the effect of winding angle on the behavior of glass/epoxy pipes under multi-axial cyclic loading. In this experimental study, winding angles of 45◦ , 55◦ , and 63◦ and pipes were exposed to internal pressure. The results showed that the optimum configuration depends on the loading ration (hoop to axial stress ratio). To simulate high sub-sea depths, E-glass fiber/epoxy tubes were tested to destruction by Pavlopoulou et al. [24] under hydrostatic external pressure leading to buckling or crushing. Different fiber architectures and winding angles were tested at a range of wall thicknesses highlighting the advantage that hoop reinforcement offers under external pressure. Colombo and Vergani [25] presented a pipe-wall thickness optimization study by varying fiber

148

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

3

volume fraction, matrix and winding angles. The algorithm uses laminate theory for the pipe stiffness an empirical formulation for strength prediction and a failure theory to be applied for each layer. The results showed that the optimum winding angles is 44.5◦ < θ < 52.5◦ if both hoop and axial stress affects the pipe at different loading ratio. It is worth remarking that the results of this algorithm could be improved by using measured UD properties rather than calculated properties. Impact events, such as dropping, knocking and/or rough handling vessels, are expected during pipeline operations. It is therefore necessary to determine the strength reduction mechanism after impact and to improve impact resistance [26]. Up to 70% reduction in the compression strength after impact is noted for composite panels in [27, 28]. Sebaey and Mahdi [29] concluded that the peak load in crushing after impact test can drop to 50% of its value without impact on GFRP pipes. Statistical analysis of impact energies was presented by Gemi et al. [30] for GFRP composite pipes. They showed using Weibull analysis that the absorbed energy depends on impact velocity, while rebound energy is independent of impact velocity. Residual burst strength after impact was assessed by Wakayama et al. [26]. The authors concluded that the addition of layers of low modulus pitch-based carbon fiber with high-compressive failure strain to the surface of filament wounded CFRP pipes in order to improve the impact resistance. Failure pressure and fatigue life of glass/epoxy composite pipe ([±55]3 ) after impact (5 J, 7.5 J and 10 J impact energies) were measured by Sari et al. [31]. Even at this low value of impact energy, leakage and eruption pressures in a static test were found to decrease. In addition, it was found that applying composite layers to impacted pipes offer a robust, more compatible, and economical solution (impact energy of 5 J, 10 J, and 15 J) and internal pressure [32]. A wider range of impact energies (12 J - 110 J) was used in [33] to impact GFRP pipes. The results showed that, at low energy, the damage is mainly due to matrix cracking and delamination whereas, fiber breakage is usually associated with higher impact energies. The residual cyclic internal pressure of hybrid glass (G) - carbon (C) fiber/epoxy pipes was studied in [34] after pre-tension and low velocity impact. The stacking sequence of the pipe wall is [±75(G)/ ± 55(C)/ ± 45(G)]. The results showed that as the pre-stress loading applied before impact loading increase, the impact damage decreases and fatigue life increases. Deniz and Karakuzu [35], studied the effect of aging in sea water on the impact resistance and compression after impact of GFRP pipes with different diameter and similar winding angle (±45). The results show that moisture absorption, salt in seawater, diameter of specimen and residual stresses produced by manufacturing process of the composite pipe have significant effect on maximum contact force, maximum deflection, absorbed energy and failure. Hawa et al. [36] studied GFRP pipes , subjected to hydrothermal aging for up to 1500 hours, followed by impact testing. The results showed that maximum force decreases and the maximum displacement increases with increasing aging time almost linearly up to a level where the maximum force, and displacement stabilizes. Burst pressure tests were also carried out to determine the influence of environment and impact. In the current paper, the winding angle effect is examined against internal pressure capacity and impact resistance of pipes for energy applications. In addition to the winding angles that are usually addressed in the literature (a pipe of single angle), this paper examines a pipe made of different winding angles against internal pressure and impact. For safety reasons, the pipes were not allowed to burst. Instead, a feedback system controls the pressure to stop adding any additional pressure whenever any drop (leakage) appears. Drop-weight impact tower was used to perform low velocity impact at different energy levels

2. Material and Manufacturing E-glass fiber and Epoxy are used for this study due to the highly expanded market of the GFRP in current piping applications. ARALDITE LY-1564 epoxy resin, mixed with hardener, is used as matrix material. The fiber volume fraction is measured for all the manufactured pipes as per the ignition test standard ASTM D2584-11. The average value of the measured fiber volume fraction for all the specimens is 49.8% with a coefficient of variation of 8.8%. Several manufacturing techniques are reported for GFRP pipes, such as centrifugal casting, hand layup, and filament winding. The latter is adopted in this work to manufacture our GFRP pipes [24]. A five-axes computer controlled filament winding machine of maximum diameter 2 m and maximum length of 6 m is used. A PVC pipe of 110 mm external diameter is used as a mandrel. Before starting the production, the PVC pipe was covered by a thick layer of mold release agent to facilitate the mandrel extraction [37].

4

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

149

GFRP pipes were manufactured using filament angle machine, with [±45]3 , [±55]3 , [±63]3 , and [±63/ ± 45/ ± 55] winding angles. Three pipes for each angle were manufactured with a total length of 1 m of each pipe. The specimen was left over night on the mandrel mounted on the machine with continuous rotation to avoid any agglomeration of the matrix at one side of the pipe. After being completely cured, the edges were trimmed and the pipes were cut into specimens of 450 mm length using diamond saw. The PVC mandrel was kept inside the pipe during the whole manufacturing and cutting process. After getting the final specimen, the PVC mandrel was extracted by a very light extraction force. The dimensions of the specimens after being prepared for the tests were measured and analyzed. The measurements of the length, wall thickness and internal diameters are 448.5 mm, 3.8 mm, and 110 mm, respectively. 3. Mechanical Testing 3.1. Internal pressure Pressure test was performed to determine the capacity and the mode of failure of each designed GFRP pipe. Hydrostatic testing is the most common method employed for testing pipes and pressure vessels. The test involves filling the pipe system with a liquid, usually water or oil, which may be dyed to aid in visual leak detection. The composite pipes are usually tested at 150% of their working pressure and commonly employed in pressure cylinder designs for the transport of dangerous goods as design pressure. Compressed Natural Gas (CNG) is usually tested at 30 MPa [38]. The burst test was performed using Resato high pressure technology machine. During the test, the first step was filling the pressure pipe to 5 bars in 30 seconds. Then, pressure was set to reach 100 bars in 60 seconds, which expected for pipes to fail before reaching the maximum pressure. At that time, the machine was set to be stabilized and decrease slowly to 0 bar which allow the machine to reach ending test stage. The fixture used for pipes is made of steel. The fixture composed of two steel plates of 500 x 500 mm in-plane dimensions and 20 mm thickness. One of the two plates has a central hole to apply the internal pressure. The two end plates were connected together by four steel threaded rods, holding the composite pipe in-between. The steel plates were machined at the interface with the pipe edge to ensure perfect sealing. In addition, the pipes were internally and externally sealed with rubber to ensure the failure away from the pipe/fixture interface. The machine control unit was prepared to stop applying the hydrostatic pressure at the first drop of the internal pressure for safety reasons. Three specimens were tested of each pipe configuration. In a typical internal pressure tests of the GFRP tubes, five important failure steps were observed. These failure steps are whitening initiation, dense whitening, leakage initiation, oil jet formation and ultimate failure [39]. As the internal pressure applied on the specimens starts to increase, the lengths of the specimens tend to shorten while their diameters enlarge. As the pressure increased, whitening initiation is observed and this tendency keeps increasing. The whitening has caused separation of fibers off the matrix interface and leads to delamination. Together with this, matrix cracks is formed in between the matrix layers on the specimens progressed and the first leakage occurs. As the internal pressure keeps increasing, the leakage turns into oil jet and the ultimate failure occurred when the tubes fails catastrophically. Pressure tests were applied to the filament-wound composite pipes in close-end condition using computer controlled hose test machine. A protection test box was confidently used for observing the failures of the specimens. During all the internal pressure tests, this test apparatus was used by satisfying the closed-end conditions of the composite pipes. 3.2. Impact test After unloaded from the internal pressure test, the composite pipes were subjected to low velocity impact. The drop weight impact tester (IMATEK machine) was used to apply different values of the impact energy to the composite pipes. The machine, the test fixture and the test configurations are shown in Figure 1. V-support was used to fix the pipes. From the upper side, a hollow plate is pneumatically supported the pipes as the impactor/striker approaches. The impact energy value (15, 30, 45 J) were selected to represent the low velocity impact characteristics [40, 41]. During all the tests, the impactor mass was set constant (8.008 kg) to avoid any possible mass effect [42]. To change the energy at the same impactor mass, the energy equation (Ei = mgh, with m is the impactor mass, g is the gravity acceleration and h is the impact height) was used to determine the impact height h. After being tested, the GFRP composite pipes were subjected to internal light source to identify the projected delamination area.

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

150

5

Figure 1. Impact machine, test set-up and configurations.

4. Results and Discussions 4.1. Internal pressure results The pressure test is divided into two main stages. Stage one is filling process at which the machine fills the GFRP pipe with the oil at 5 bars. Stage two is pressurizing process, where the machine starts to pressure the liquid in the pipe up to the peak point. Failure of composite pipe is usually initiated by matrix cracking. Therefore, on increasing the pressure, it is seen that leakage failure has occurred at the specimen surface. The failure propagates with additional matrix failure as a result of increasing the cycles, causing a drop in the internal pressure reading which stopped the loading processes. The internal pressure history is summarized in Figure 2 for the entire test campaign.

Pressure (bar)

80 60

Orientation 45

Orientation 55

Orientation 63

Combined

40 20 0

0

20

40

60

80

100

Time (second) Figure 2. Loading history of the four pipes with different filament angles.

For the pipes made of ±55◦ winding angles, during the filling process there was no sign of leakage. After that, the pressure increased in stage two up to 56 bars. None of the three tested samples showed leakage at the specimen’s

6

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

151

mid-section. Instead, the three of them showed leakage at/near the interface with the steel plates. This result reveals that the capacity of this pipe is higher than the measurements. For the specimens made of ±63 winding angles, no leakage was observed during the first stage as the pressure increased up to 5 bars. After that, the pressure increase until the peak point. The failure occurs as a leakage, of few drops, on the specimen surface. After the test, the pipes were examined by naked eye, and there was no cracking, whitening or delamination. The specimens manufactured with the ±45 were not expected to provide the best internal pressure capabilities as they are not amongst the recommended stacking sequence for internal pressure application. During the manufacturing process of such pipe, fibers were not evenly distributed throughout the length of the pipe. There was a fiber overlaps, which increased the distance between the fibers, and created gaps in the pipe. The maximum pressure that the pipe can handle was 7 bars after that pressure, the leakage started all over the pipe surface. cracking, whitening or delamination did not appear on the pipe by naked eye examination. It is worth remarking that this low capacity under internal pressure was a result of the gaps generated during the manufacturing. The specimen manufactured with combined filament angles showed hybrid response. After filling process, the pressure increases until the failure point (23 bar). The failure is leakage weeping along the lower surface of the pipe, no crack or delamination occur on the pipe surface. It is highly believed that the result of this hybrid specimen is affected by the gabs in the ±45 layer which caused major thinning in the specimen wall-thickness. Figure 2 shows the peak point for all specimens (55,63,45 and combined) under the internal pressure test. In summary, the pipe with staking sequence of ±55 showed highest peak point (56 bar). The ones with ±63 and combined GFRP pipe handled pressures of 34 bar and 23 bars, respectively. While the failure on ±45 pipe occurs at 7 bars. In the oil and gas applications, different design considerations are usually taken when selecting a certain pipe configuration depending on the function. In general, up to 50 mm (2 in) diameter pipe is used to for the distribution lines. These type of pipes lines are usually carrying an internal pressure up to 15 bar. Gathering pipelines, used between the source and processing, are usually of 50 – 200 mm diameter (2-8 in) and working for a typical pressure level of 45 – 50 bar. Transmission lines are of higher diameter (up to 1500 mm) and working under a higher level of pressure (up to 2000 bar) [43]. The pipes tested within this paper can suit the first and the second type of pipelines. The ones made of the ±55 and ±63 can fit for both applications whereas the hybrid ones can be recommended for the distribution lines. Although it is usually recommended for bucking design consideration, layers of ±45 winding angles cannot be recommended, based on this analysis, for any of the prescribed applications. 4.2. Impact results Typical results of the standard impact test are usually the load-time and energy-time plots, Figure 3. The results can be assessed using several parameters [40, 41]. In the current study, the load capacity and the energy capacity are assessed for the entire test matrix. For each case, the maximum load, the delamination threshold load [27] (the load at which the first drop in the load-time plot appears), the elastic energy (the amount of recovered energy), and the energy consumption (the amount of energy consumed in damage) [44] are measured. Figure 4 shows a summary of the four parameters with respect to the impact energy. In addition, after impact, the specimens are visually checked for the amount of damage. Figure 5 shows a summary of the measured delamination/damage area for all the pipes at different impact energies. As a general observation of Figure 3, the ±63 pipes have the minimum value of the maximum force, as compared to the other pipes. The fact is also obvious in Figure 4(a). This lower load capacity results in longer contact between the pipe and the striker which is an indication of higher displacement. For the same pipe material, the elastic energy is minimum which indicates that most of the impact energy is consumed in damage. The lower amount of damage associated with this pipe, Figure 5, can be justified with the fact that it is a projected delamination area (which is far away from the actual delamination area in the case of multi-directional laminates) [28]. From the impact point of view, this pipe is not recommended. For the pipes manufactured using the ±55 and combined orientations, very similar load-time profiles and data are obtained, Figures 3(a), 3(c), and 3(e). A slightly higher values of the maximum load is recorded for the combined orientations, Figure 3(c). Also, higher values of the elastic energies are recorded for the combined orientations. From the mechanical point of view, this preference of the combined orientations can be justified with the possible effect of mismatch angle on the fracture toughness of the materials and consequently on the damage imitation and propagation of the pipes [45]. The comparison between the delamination threshold values of different pipes, Figure 4(b), shows

Author name / Energy Procedia 00 (2019) 000–000 Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155

152 3

Orientation 55

2

Orientation 63

1.5

Combined

1

14

Energy (J)

Impact Force (kN)

16

Orientation 45

2.5

12 10 8 6 4

0.5 0

2 0

5

10

15

0

20

0

Orientation 45

Orientation 55

Orientation 63

Combined

5

10

time (ms)

4.5

35

Orientation 55

3

Orientation 63

2.5 2

Combined

1.5

30

Energy (J)

Impact Force (kN)

3.5

25 20 15 10

1 0.5

5

0

0

10

15

20

25

30

0

5

Orientation 45

Orientation 55

Orientation 63

Combined

10

time (ms)

Orientation 63 Combined

10

15

20

time (ms) (e) Load-time for 45 J impact energy

25

30

Energy (J)

Impact Force (kN)

Orientation 55

5

20

25

30

(d) Energy-time for 30 J impact energy

Orientation 45

0

15

time (ms)

(c) Load-time for 30 J impact energy

5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0

20

(b) Energy-time for 15 J impact energy

Orientation 45

4

5

15

time (ms)

(a) Load-time for 15 J impact energy

0

7

50 45 40 35 30 25 20 15 10 5 0

0

5

Orientation 45

Orientation 55

Orientation 63

Combined

10

15

20

25

30

time (ms) (f) Energy-time for 45 J impact energy

Figure 3. History of the load vs. time and energy vs. time for the whole test matrix.

a lower value for the ±63 pipes which indicates that the damage in these pipes starts earlier, compared to the others. In contrary, the combined and the ±55 pipes show the higher values of the delamination threshold which indicates a delay in damage initiation. For the projected delamination area, the preference is given to the ±55 pipes. The higher

Author name / Energy Procedia 00 (2019) 000–000 Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0

10

Orientation 45

Orientation 55

Orientation 63

Combined

20

30

40

2 1.5 1 0.5

50

0

10

20

(a) Maximum force as a function of the impact energy

40

12

Elastic Energy (J)

25 20 15 10

50

Orientation 45 Orientation 55 Orientation 63 Combined

10 8 6 4 2

5 10

20

30

40

0

50

10

20

Impact Energy (J)

30

40

Impact Energy (J)

(c) Energy cnsumed in damage vs. impact energy

(d) Elastic energy vs. impact energy

Figure 4. summary on the maximum force, delamination threshold, energy consumption, and elastic energy with the impact energy.

7000

Orientation 45 Orientation 55 Orientation 63 Combined

6000

Delamination Area (mm2)

Energy consumption (J)

30

40

(b) Delamination threshold as a function of the impact energy

Orientation 45 Orientation 55 Orientation 63 Combined

35

30

Impact Energy (J)

Impact Energy (J)

0

153

2.5

Delamination Threshold (kN)

Maximum Force (kN)

8

5000 4000 3000 2000 1000 0

10

20

30

40

50

Impact Energy (J) Figure 5. Projected delamination area for the entire test matrix as function of the impact energy.

50

154

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

9

values of the projected delamination area in Figure 5 for the combined and the ±45 can be justified by the number of delaminations rather than the size of a single delamination. 5. Conclusions This paper presents an experimental investigation on the composite pipe internal pressure capacity and impact. The paper tested four pipe configurations made with the winding angles of [±45]3 , [±55]3 , [±63]3 , and [±63/ ± 45/ ± 55]. All the specimens have the same nominal dimensions of 110 mm internal diameter, 3.8 mm wall thickness and 448.5 mm length. Internal pressure tests were conducted under closed loop control system to ensure safety. After the internal pressure test, low velocity impacts were applied to the specimens using a drop-weight impact tester. The pressure results showed the superior internal pressure capacity of the specimens manufactured using the winding angles ±55, which in agreement with the data available in the literature. The specimens made of 63 winding angles and the ones made of the hybrid angles both showed a close values of the pressure capacity, with advantages to the ±63 specimens due to the gaps resulted during manufacturing of the ±45 layers. Unfortunately, the specimens made of all ±45 angles are not recommended for any oil and gas applications based on our current analysis and the manufacturing difficulties we faced with this configuration. From the impact results, the combined orientations ([±63/ ± 45/ ± 55]) pipe can be considered as a competitor to the ±55 pipes. Although, their low pressure capacity might limit their usage to the low pressure applications. The pipes manufactured using the ±63 orientations show a very low damage resistance, as assessed using all the prescribed parameters. Obviously, the decision is clear in case of selecting a pipe for only the pressure capacity. However, pipes are not only subjected to internal pressure and they might suffer from impact, bending, axial loadings, aging, etc. during their lifetime. The response of composite part under any of these conditions is highly dependent on the winding angles. For this reason, a recommendation can be drawn to design composite pipes based on the satisfaction of, not only the internal pressure, but also the other expected loading and working conditions. References [1] R. Rafiee, On the mechanical performance of glass-fibre-reinforced thermosetting-resin pipes: A review, Composite Structures 143 (2016) 151–64. [2] M. Lakhdar, D. Mohammed, L. Boudjemaa, A. Rabia, M. Bachir, Damages detection in a composite structure by vibration analysis, Energy Procedia 36 (2013) 888–97. [3] E. Giannaros, T. Kotzakolios, V. Kostopoulos, Blast response of composite pipeline structure using finite element techniques, Journal of Composite Materials 50 (2016) 3459–76. [4] R. Wongpajan, S. Mathurosemontri, R. Takematsu, H. Y. Xu, P. Uawongsuwan, S. Thumsorn, H. H., Interfacial shear strength of glass fiber reinforced polymer composites by the modified rule of mixture and kelly-tyson model, Energy Procedia 89 (2016) 328–34. [5] K. B. Fischer, A. Daga, D. Hatchell, G. T. Rochelle, MEA and piperazine corrosion of carbon steel and stainless steel, Energy Procedia 114 (2017) 1751–64. [6] S. Lu, L. Gao, Q. Li, X. Han, D. Zhao, On-line monitoring technology for internal corrosion of CO2 -EOR oil field, Energy Procedia 154 (2018) 118–24. [7] E. Latosov, B. Maaten, A. Siirde, A. Konist, The influence of O2 and CO2 on the possible corrosion on steel transmission lines of natural gas, Energy Procedia 147 (2018) 63–70. [8] G. Banushi, I. Weidlich, Seismic analysis of a district heating pipeline, Energy Procedia 149 (2018) 216–25. [9] M. Bouhafs, Z. Sereir, A. Chateauneuf, Probabilistic analysis of the mechanical response of thick composite pipes under internal pressure, International Journal of Pressure Vessels and Piping 95 (2012) 7–15. [10] T. van der Zee, M. Jan de Ruiter, I. Wieling, The C-tower project - A composite tower for offshore wind turbines, Energy Procedia 137 (2017) 401–5. [11] J. M. Lees, Behaviour of GFRP adhesive pipe joints subjected to pressure and axial loadings, Composites: Part A 37 (2006) 1171–9. [12] R. Rafiee, B. Mazhari, Simulation of the long-term hydrostatic tests on glass fiber reinforced plastic pipes, Composite Structures 136 (2016) 56–63. [13] G. Cz´el, T. Czig´any, A study of water absorption and mechanical properties of glass fiber/polyester composite pipes: Effects of specimen geometry and preparation, Journal of Composite Materials 42 (2008) 2815–27. [14] DNVGL-RP-F119, Recommended Practice for Thermoplastic composite pipes, Technical Report, DNV GL, 2015. [15] G. Wr´obel, M. Szymiczek, J. Kaczmarczyk, Influence of the structure and number of reinforcement layers on the stress state in the shells of tanks and pressure pipes, Mechanics of Composite Materials 53 (2017) 165–78.



10

Tamar Ali Sebaey / Energy Procedia 162 (2019) 146–155 Author name / Energy Procedia 00 (2019) 000–000

155

[16] P. A. Zinoviev, S. V. Tsvetkov, Mechanical properties of unidirectional organic-fiber-reinforced plastics under hydrostatic pressure, Composites Science and Technology 58 (1998) 31–9. [17] M. Xia, K. Kemmochi, H. Takayanagi, Analysis of filament-wound fiber-reinforced sandwich pipe under combined internal pressure and thermomechanical loading, Composite Structures 51 (2001) 273–83. [18] M. W. K. Rosenow, Wind angle effects in glass fiber reinforced polyester filament wound pipes, Composites 15 (1984) 144–52. [19] A. F. Hamed, A. Y. Khalid, S. M. Sapuan, M. M. Hamdan, T. S. Younis, B. B. Sahari, Effects of winding angles on the strength of filament wound composite tubes subjected to different loading modes, Polymers and Polymer Composites 15 (2007) 199–206. [20] F. A. R. Haftchenari, H. Al-Saleti, S. T. S. Al-Hassani, Effect of temperature on the tensile strength and failure modes of angle ply aramid fibre (KRP) tubes under hoop loading, Applied Composite Materials 9 (2002) 99–115. [21] R. Rafiee, B. Mazhari, Modeling and experimental evaluation of functional failure pressures in glass fiber reinforced polyester pipes, Computational Materials Science 96 (2015) 579–88. [22] I. Y. Sulu, S. Temiz, Stress analyses of multi-layered composite pipes subjected to internal pressure, Academic Journal of Science 4 (2015) 187–94. [23] P. Krishnan, M. S. Abdul Majid, M. Afendi, A. G. Gibson, H. F. A. Marzuki, Effects of winding angle on the behaviour of glass/epoxy pipes under multiaxial cyclic loading, Materials and Design 88 (2015) 196–206. [24] S. Pavlopoulou, S. S. Roy, M. Gautam, L. Bradshaw, P. Potluri, Numerical and experimental investigation of the hydrostatic performance of fibre reinforced tubes, Applied Composite Materials 24 (2017) 417–48. [25] C. Colombo, L. Vergani, Optimization of filament winding parameters for the design of a composite pipe, Composites: Part B 148 (2018) 207–16. [26] S. Wakayama, S. Kobayashi, T. Imai, T. Matsumoto, Evaluation of burst strength of FW-FRP composite pipes after impact using pitch-based low-modulus carbon fiber, Composites: Part A 37 (2006) 2002–10. [27] T. A. Sebaey, E. V. Gonz´alez, C. S. Lopes, N. Blanco, P. Maimi, J. Costa, Damage resistance and damage tolerance of dispersed CFRP laminates: Effect of the mismatch angle between plies, Composite Structures 101 (2013) 255–64. [28] T. A. Sebaey, E. V. Gonz´alez, C. S. Lopes, P. Blanco, J. Costa, Damage resistance and damage tolerance of dispersed CFRP laminates: Effect of ply clustering, Composite Structures 106 (2013) 96–103. [29] T. A. Sebaey, E. Mahdi, Crashworthiness of pre-impacted glass/epoxy composite tubes, International Journal of Impact Engineering 92 (2016) 18–25. [30] L. Gemi, M. Kayrici, M. Uludag, D. S. Gemi, O. S. Sahin, Experimental and statistical analysis of low velocity impact response of filament wound composite pipes, Composites: Part B 149 (2018) 38–48. [31] M. Sari, R. Karakuzu, M. E. Deniz, B. M. Icten, Residual failure pressures and fatigue life of filament-wound composite pipes subjected to lateral impact, Journal of Composite Materials 46 (2011) 1787–94. [32] M. Kara, M. Uyaner, A. Avci, Repairing impact damaged fiber reinforced composite pipes by external wrapping with composite patches, Composite Structures 123 (2015) 1–8. [33] R. F. Alshahrani, N. Merah, S. M. A. Khan, Y. Al-Nassar, On the impact-induced damage in glass fiber reinforced epoxy pipes, International Journal of Impact Engineering 97 (2016) 57–65. [34] L. Gemi, O. S. Sahin, A. Akdemir, Experimental investigation of fatigue damage formation of hybrid pipes subjected to impact loading under internal pre-stress, Composites: Part B 119 (2017) 196–205. [35] M. E. Deniz, R. Karakuzu, Seawater effect on impact behavior of glass-epoxy composite pipes, Composites: Part B 43 (2012) 1130–8. [36] A. Hawa, M. S. Abdul Majid, M. Afendi, H. F. A. Marzuki, N. A. M. Amin, F. Mat, A. G. Gibson, Burst strength and impact behaviour of hydrothermally aged glass fibre/epoxy composite pipes, Materials and Design 89 (2016) 455–64. [37] G. Perillo, R. Vacher, F. Grytten, S. Sorbo, V. Delhaye, Material characterisation and failure envelope evaluation of filament wound GFRP and CFRP composite tubes, Polymer Testing 40 (2014) 54–62. [38] G. W. Mair, M. Hoffmann, A. Schoppa, M. Szczepaniak, Slow burst testing of samples as a method for quantification of composite cylinder degradation, International Journal of Hydrogen Energy 39 (2014) 20522–30. [39] M. Kara, M. Uyaner, A. Avci, A. Akdemir, Effect of non-penetrating impact damages of prestressed grp tubes at low velocities on the burst strength, Composites: Part B 60 (2014) 507–14. [40] ASTM-D7136/D7136M, Standard test method for measuring the damage resistance of a fiber-reinforced polymer matrix composite to a dropweight impact event, ASTM International, American Society for Testing and Materials, West Conshohocken, PA, USA, 2007. [41] AITM1-0010, Fiber reinforced plastics: Determination of compression strength after impact, Airbus Test Method, Blagnac Cedex, France, 2005. [42] R. Olsson, Mass criterion for wave controlled impact response of composite plates, Composites: Part A 31 (2000) 879–87. [43] A. Hilier, Types of pipeline every oil and gas engineer should know about- mirage machines portable performance blog, https://blog. miragemachines.com/types-of-pipeline-every-oil-and-gas-engineer-should-know-about, 2017. [44] E. V. Gonz´alez, P. Maim´ı, P. P. Camanho, C. S. Lopes, N. Blanco, Effect of ply clustering in laminated composite plates under low-velocity impact loading, Composites Science and Technology 71 (2011) 805–17. [45] T. A. Sebaey, N. Blanco, C. S. Lopes, J. Costa, Characterization of crack propagation in mode I delamination of multidirectional CFRP laminates, Composites Science and Technology 72 (2012) 1251–6.